GTM Engineering Proposal for Dobbin
Built with Steven Brady's AuthIn GTM method
Dobbin Business Machines, September 2026
Nine minutes on how I read your situation. The plays below are what's on screen.
Executive summary
Dobbin launched self-serve this month. Anyone can click a few buttons and get a working company AI, built from their public presence, in under 15 minutes, for about $25 in build cost. The next jobs are getting more of those people past the Slack install, and filling the top of the funnel with the right people.
This proposal lays out five outbound plays built for Dobbin's buyers: agencies that represent artists, design studios, and Shopify brands. Each play starts from a specific list, built from public data, and each message gives the prospect something useful before asking for anything.
“The message isn't the problem. The LIST is the message.”
Steven Brady, AuthIn GTMDobbin has an advantage almost no other company has here. Your product can build a prospect's own AI from their website alone. So the most valuable thing an outbound message can carry is their own working Dobbin, built before they ever reply.
Where Dobbin's outbound stands today
From the job post, public sources and our call on 25 September:
- All sales so far have been founder-led, hand to hand, with about 25 paying customers including Evite, Area17, Good Dye Young, Mark Seliger's studio, Giant Artists and Highlights Media.
- Matt has been running LinkedIn outreach through HeyReach for about three months, sending 20 to 25 a day from Josh's profile. A fractional content marketer is building Dobbin's LinkedIn presence.
- Self-serve setup most often stalls at the Slack install, when the person who started it can't add apps to their workspace.
- Three member networks with about 1,900 members in total have a direct relationship with Dobbin. Two AMA agencies are now customers, and the head of DTCMVP is building his own Dobbin, but none of them is run as a channel yet.
- The stack is Framer, Pipedrive, Apollo (now with website visitor tracking), Klaviyo, Zapier, Stripe and Notion.
Most outbound in this category leans on the same Apollo filters everyone else uses: agencies with 10 to 50 employees, DTC brands on Shopify. Those lists get the same messages from every AI vendor. The plays below start from a prospect's current situation, found through signals most teams never check.
The AuthIn GTM approach
The AuthIn method runs two kinds of outbound, and both depend on building the list first.
Pain-Qualified Segments (PQS)
A PQS is a list of companies in a specific, checkable situation right now, found through public signals. For Dobbin, that looks like an agency hiring its second producer to keep up with bids.
The message describes their situation back to them precisely enough that the reaction is "how did they know?"
Permissionless Value Props (PVP)
A PVP hands the prospect something useful before any sales conversation, something they'd keep even if they never reply.
For Dobbin, the PVP is the product itself: their own company AI, already built from their public presence and ready to try.
Five plays for Dobbin
Highlighted fields fill per prospect from the data recipe. Customer results quoted are Dobbin's own published numbers.
Play 1: “Your agency's AI, already built”
Data recipe: AMA member directory + agency roster pages + a pre-built Dobbin instance
Pull the AMA's roughly 250 member agencies, scrape each roster page for artist count and specialties, then trigger a Dobbin build from the agency's site before the first touch. Sent from Josh's LinkedIn through HeyReach.
Subject: {agency}'s AI, already built
Ran {agency}'s site through Dobbin last night. It built a working AI that knows your {roster_count} artists, how you pitch them, and the tone of your treatments.
Mark Seliger's studio uses theirs to answer RFPs in 30 minutes, down from two weeks.
Want the link? It's live for 30 days whether or not we talk.
At about $25 a build, pre-building is cheap, and the budget gets set once the first segments are picked. Seliger's studio and Giant Artists are already customers, so the rest of the membership knows the names.
Play 2: The bid-season bottleneck
Data recipe: agency job posts for producer and bidding roles + roster growth on the agency site + AMA membership
Watch LinkedIn Jobs and the AMA job board for agencies hiring executive producers, estimators or bid coordinators. Cross-check roster size over time from the site. An agency adding artists and hiring producers has an owner still writing every treatment and estimate.
Subject: the {role} opening
Saw {agency} is hiring a {role} while the roster's grown to {roster_count} artists. Usually means the owner is still writing every bid.
Seliger Studio had the same problem. With Dobbin trained on Mark's point of view, the team now answers RFPs in 30 minutes and just had its best revenue in over seven years.
Curious how they set it up?
The Seliger result lands hardest with the AMA because its members compete for the same commercial jobs.
Play 3: Founder voice at a growing Shopify brand
Data recipe: Shopify store detection + LinkedIn headcount growth + founder posting frequency + first brand or marketing manager hire
Start from Shopify brands (StoreLeads or BuiltWith), keep those that grew headcount 30% or more in 12 months, then keep the ones whose founder posts weekly and who just hired a first brand or marketing lead. That combination points to a founder who is still the brand's only real source of truth.
Subject: {headcount} people, one founder voice
{brand} grew from {headcount_prior} to {headcount} people this year and you're still posting {post_frequency}. Guessing most brand calls still route through you.
Galanter & Jones hit the same wall. After putting their strategy and voice into Dobbin, they grew 50% in H2 with the same team.
Want to see yours? It builds from your site in about 15 minutes.
DTCMVP's 1,000 operators come from brands like Kosas, Dagne Dover, Kith and Cuts, which makes them the right people to test this message on before scaling it. Sean, who runs the network, is building his own Dobbin, which gives members a demo from someone they know.
Play 4: A design studio's proposal voice
Data recipe: DLN member firms + project pages on each studio's site + press features (AD100, Elle Decor A-List)
For DLN principals, gather the studio's published projects and press features, then build their Dobbin from that material so the first message shows it drafting a proposal in the studio's own language.
Subject: {studio}'s proposal voice
Fed the {project_count} projects on {studio}'s site and your {publication} feature into Dobbin. It now drafts proposals and client updates in your studio's voice.
Area17 already runs on Dobbin.
Want the login so your team can try it on the next proposal?
DLN principals meet in forum groups of 7 to 9 and at a 150-person Leadership Summit, so one studio that sees value tends to reach the rest quickly.
Play 5: People already asking for a company brain
Data recipe: engagement on Josh's posts and on "company brain" posts + ICP filter in Clay + Apollo contact data
Trigify tracks who comments on or reacts to Josh's posts and to posts about company knowledge tools. Clay keeps the ones at 10 to 200 person companies in Dobbin's verticals. The message quotes the comment back to them.
Subject: your comment on {post_author}'s post
Saw your comment on {post_author}'s post about {topic}, the part about {their_words}.
Dobbin does that inside Slack. It answers the way your founders would, from your own docs and calls, and teams are live in about 15 minutes.
Want me to build one for {company} so you can try it?
This is the Mister Brady Method loop: 45% acceptance and 52% reply on the published HeyReach playbook.
Implementation components
1. Data sources
- Network sources: AMA member directory and job board, DLN member and event lists, DTCMVP's paid feedback calls
- Industry sources: agency roster pages, studio project pages, AD100 and Elle Decor A-List, Shopify store detection
- Hiring signals: producer, bidding, brand and marketing roles on LinkedIn Jobs and the AMA board
- Social signals: engagement on Josh's posts and on company-knowledge posts, tracked through Trigify
2. Data engine
- Clay or Deepline tables that run each recipe on a schedule and flag new matches
- Apollo for contact data, since Dobbin already pays for it
- HeyReach across Josh's profile and any added sending accounts, run day to day by Matt
- Zapier pushing every reply, install and Stripe event into Pipedrive
3. Message development
- PQS messages describe one situation the prospect will recognize as theirs
- PVP messages carry a built Dobbin, or a piece of one, as the attachment
- Testing: each play runs on a small segment first, and I report acceptance, reply, install and paid rates weekly in Slack
4. The three networks as channels
| Network | What exists today | Dobbin's way in |
|---|---|---|
| AMA (250) | Seliger's studio and Giant Artists as customers, and a member perks page where vendors like Lettuce Financial offer a free month and a referral bounty | A member-only Dobbin offer on the perks page, a live session led by a current member, and warm intros through the former agent Josh is talking to |
| DLN (650) | 85+ live programs a year, virtual roundtables of 20 to 75, and a partner roster including Design Within Reach and Kohler | A virtual program on how studios use AI for proposals, with Area17 as the proof |
| DTCMVP (1,000) | A paid feedback marketplace where SaaS companies book intros with operators from brands like Kosas and Kith, at a few hundred dollars each | Sean's own Dobbin as the demo, then paid intros that end with a live build of the operator's Dobbin |
Why this works for Dobbin
1. Lists nobody else is building. Other AI vendors email "creative agencies, 10 to 50 employees." These plays reach the agency hiring its second producer and the founder still posting weekly at 60 people.
2. The product is the offer. Few companies can send a prospect their own working product before the first reply. Dobbin can, and the PVP plays are built around that.
3. Recipes get better over time. Each week's replies and installs show which signals predict paid accounts, and those signals get more weight in the next build.
4. Numbers from the same method:
- $49K in revenue in 8 weeks, with 45% acceptance and 52% reply, on the published HeyReach playbook
- Champify: 572 connection requests sent into closed-lost accounts, which produced 8 meetings and about $380K in pipeline in 4 weeks
- Keep: 25 LinkedIn accounts, 10,000+ connection requests a month and about 120 meetings a month for 14 months
Implementation timeline
Data and baseline
- Map the 25 current customers by vertical, size and how they found Dobbin
- Measure where setups stall at the Slack install, and ship the fix with Joonas
- Build the source lists for the AMA, DLN and DTCMVP
- Tag every install link by source
Recipes live
- Plays 1 and 2 running on the AMA
- Play 5 running on engagement with Josh's posts
- Signal strength checked against the first replies and installs
Messages tested
- Plays 3 and 4 tested on small DTCMVP and DLN segments
- The first live network session run
- Losing message versions cut, winners moved to more sending accounts
Automation and scale
- Recipes running on a schedule, with new matches going straight into HeyReach
- Klaviyo sequences keyed to what trial users do in the product
- The full playbook written up in Notion and a day 90 readout to both founders
Investment
Engagement
- Contract: $12,000 a month for 90 days, as posted, with the option to talk about a permanent role at day 90
Tools
- Already in place: Apollo, Pipedrive, Klaviyo, Zapier, Stripe
- Builds: $25 per build, budgeted once the first segments are picked
- To add: HeyReach seats per sending account, Trigify, and Clay or Deepline credits. I'll price these in week one against the list sizes above.
- Network costs: DTCMVP intro fees and any AMA or DLN partner fees, decided after week two
Team
- A weekly brief for Matt: he keeps building and sending, and I set which segments and messages go out
- A handoff session with the fractional content marketer in week one
- Everything documented in Notion so the next hire runs the same system
Next steps
The 90 day plan and funnel map are here.
A call with Josh and Joonas to pick this apart and agree on the first two plays. I can start the next day.